INDEX

Explanations

specific identifiers and entities related to names and titles

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

DISABLE

-0.13

eof

-0.13

 succesfully

-0.13

Äįet

-0.12

 whats

-0.12

plied

-0.12

 wich

-0.12

 facto

-0.12

cuz

-0.12

culator

-0.12

POSITIVE LOGITS

ÂŃi

0.17

ÂŃing

0.16

.define

0.16

()↵↵

0.15

 Quotes

0.15

_↵↵

0.14

 :↵↵↵↵

0.14

 Banc

0.14

_NONNULL

0.14

(íģ¬ê¸°

0.14

Activations Density 2.517%

specific identifiers and entities related to names and titles

No Comments

No Known Activations